NVIDIANemotron 3 Ultra 550B A55B (Reasoning)
- Released
- Jun 4, 2026
- Tracked since
- 2026-07-23
Intelligence index22.9Ranks 169 of 664 models with this metric
Coding index49.3Ranks 114 of 259 models with this metric
Agentic index20.1Ranks 88 of 156 models with this metric
Capability profile
5 axes vs. site averageNemotron 3 Ultra 550B A55B (Reasoning)
Site average
Pricing
per 1M tokens- Input
- $0.600
- Output
- $2.50
- Cache read
- $0.160
Performance
- Output throughput
- 145 tokens/sec
- Time to first token
- 2.16s
- First answer token
- 17.82s
- End-to-end response
- 21.26s
History
13Intelligence index
37.8→22.9
Output
$2.67→$2.50
Output throughput
194→145
Cheaper at this level
Models with a comparable intelligence index and a lower output price- Hy3-preview (Reasoning)
TencentIntelligence 22.7$0.210−92% - Ling 3.0 Flash
InclusionAIIntelligence 24.9$0.220−91% - Ling-3.0-flash-VLInclusionAIIntelligence 24.6$0.220−91%
- Ling-3.0-flash-FinInclusionAIIntelligence 22.6$0.220−91%
- DeepSeek V4 Flash 0420 (Reasoning, High Effort)
DeepSeekIntelligence 26$0.280−89%